Ticket 8712

Summary:	Configless slurmd does not work with pam_slurm_adopt
Product:	Slurm	Reporter:	Trey Dockendorf <tdockendorf>
Component:	Configuration	Assignee:	Jacob Jenson <jacob>
Status:	RESOLVED FIXED	QA Contact:
Severity:	6 - No support contract
Priority:	---	CC:	bas.vandervlies, jog, lloyd_brown, uemit.seren
Version:	20.02.2
Hardware:	Linux
OS:	Linux
Site:	-Other-	Alineos Sites:	---
Atos/Eviden Sites:	---	Confidential Site:	---
Coreweave sites:	---	Cray Sites:	---
DS9 clusters:	---	HPCnow Sites:	---
HPE Sites:	---	IBM Sites:	---
NOAA SIte:	---	NoveTech Sites:	---
Nvidia HWinf-CS Sites:	---	OCF Sites:	---
Recursion Pharma Sites:	---	SFW Sites:	---
SNIC sites:	---	Tzag Elita Sites:	---
Linux Distro:	CentOS	Machine Name:
CLE Version:		Version Fixed:	v20.02.2
Target Release:	---	DevPrio:	---
Emory-Cloud Sites:	---

Description Trey Dockendorf 2020-03-20 06:46:11 MDT

I have setup configless for slurmd but it appears like pam_slurm_adopt still expects to read /etc/slurm/slurm.conf.

Mar 20 08:40:48 pitzer-slurmd01-test pam_slurm_adopt[49090]: error: s_p_parse_file: unable to status file /etc/slurm/slurm.conf: No such file or directory, retrying in 1sec up to 60sec

If I create a symlink from /run/slurm/conf/slurm.conf to /etc/slurm/slurm.conf then the PAM module works just fine.

Comment 1 Jake Grimmett 2020-05-05 04:58:44 MDT

Thanks for the tip.

the symlink works, however when you restart slurmd it sees /etc/slurm/slurm.conf and so doesn't update /var/spool/slurmd/conf-cache this results in a stale config when any changes are made to slurm.conf.

The temporarily fix is:
systemctl stop slurmd && rm -f /var/spool/slurmd/conf-cache/* && systemctl start slurmd

Perhaps this could be added to /usr/lib/systemd/system/slurmd.service ?

Comment 2 Jake Grimmett 2020-05-13 07:39:36 MDT

This bug was fixed in slurm v20.02.2

Changes in Slurm 20.02.2
(SNIP)
>  -- pam_slurm_adopt - explicitly call slurm_conf_init().

:)